An experimental study was carried out to understand the physico-chemical and mechanical properties of marine clay reconstituted with different pore fluids. Three different pore fluids namely distilled water, 0.4 M NaCl and 1.0 M NaCl solutions, and 0.4 M CaCl2 solution were used in this study. The specimens were prepared using a 1D slurry consolidation technique at 50 kPa vertical pressure. This paper mainly includes the microstructural studies conducted using Scanning electron microscopic (SEM) images and Mercury intrusion porosimetry (MIP) tests. Furthermore, cyclic triaxial and resonant column tests were carried out on the marine clay specimens reconstituted with 0.4 M NaCl and 0.4 M CaCl2 solutions subjected to different confining pressures. The experimental results illustrated that with an increase in concentration of pore fluid the cyclic properties of reconstituted Chennai marine clay increases for strain amplitude varying between 0.001 and 1%.

In geotechnical centrifuge modelling the use of a pore fluid having viscosity greater than distilled water is a well-established method to satisfy the scaling laws required to model by a single prototype both diffusion and dynamic phenomena. However, the use of a pore fluid different from distilled water might influence the physical and hydro-mechanical properties of this soil. To analyse these possible effects, the paper presents results of a quite comprehensive experimental study carried out on Pontida sandy clayey silt compacted using two different pore fluids, namely hydroxypropyl methylcellulose (HPMC) and distilled water, recently used for centrifuge modelling in Seriate (Italy). Comparing the optimum Proctor standard of both mixtures, it was found that for the soil tested the use of HPMC has little effect on strength and small strain behaviour but affects the Atterberg limits, the Proctor compaction curve, the soil compressibility, the soil-liquid retention curve and the hydraulic behaviour both in the saturated and in the unsaturated soil range. Beside discussing the above, the aim of the paper is also to present a possible laboratory testing strategy to detect the soil properties relevant to numerical simulation of the observed behaviour of model soil structures used in geotechnical centrifuge testing.